Chronic lymphocytic leukemia (CLL) is the most common type of leukemia in the Western world. It remains incurable by available standard treatment and new therapeutic strategies are needed. In this proposal, support is requested to conduct a phase II clinical trial of bryostatin 1 (bryo 1) followed by 2 chlorodeoxyadenosine (2-CdA) in relapsed CLL. The rationale for the trial is based on a preclinical animal model revealing interaction between these two agents with therapeutic benefit. The model established in the P.I.~s laboratory utilizes an EBV-negative CLL line(WSU-CLL) that grows in mice with severe combined immune deficiency (SCID). Similar to fresh CLL cases, bryo 1 induced differentiation of WSI-CLL cells to a hairy cell stage. The proposed trial will test the hypothesis that bryo 1-differentiated Cll cells are as sensitive to 2-CdA as the de novo hairy cell leukemia. Bryo 1 will be given at the maximum tolerated dose of 120 mu/m2 by continuous intravenous intravenous infusion (CIV) over 72 has determined by a recently completed phase I trial at the applicant~s institution. It will be followed immediately by a 5 day CIVI of 2-CdA and treatment repeated every 4-8 weeks. The first cohort of 3 patients will receive 85% of standard 2-CdA dose (0.085 mg/kg/d) and escalated to 100% in the second cohort and subsequent patients unless myelosuppression is noted. In preclinical studies, sequential use of bryo1 and 2-CdA did not result in added toxicity. Laboratory correlative studies will include 3-colorflow cytometric analysis of leukemic cells to determine their differentiation state. From the statistical point, a two-stage design will be used with early stopping rules incase of extreme results. The regimen will be considered ineffective if the response rate (RR equals CR plus PR) is less than 40%, but would be of interest if RR is greater than 70%. It is estimated that 11 response-evaluable (r-e) patients are needed if extreme results are noted. If indeterminate results are observed (R 40 - 70%), 10 additional r-e patients will be accrued. This design will provide 3% type l error and 83% power to test the hypothesis. With the additional 3-6 patients needed for dose-escalation of 2-CdA, the maximum number of patients needed is 27. A consortium agreement is formed between the Detroit Medical center, Detroit VA Hospital and Henry Ford Hospital with estimated contribution of 5 patients per institution per year.